 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* University of Maryland School of Medicine, Baltimore, Maryland
University of North Carolina School of Public Health, Chapel Hill, North Carolina
University of North Carolina School of Medicine, Chapel Hill, North Carolina
|| Washington State Department of Social and Health Services, Seattle, Washington
¶ Juvenile Protection Association, Chicago, Illinois
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Design. Cohort study.
Setting. Participants included low-income families recruited from 4 sites: East, Northwest, Midwest, and South, who are part of LONGSCAN, a longitudinal study of children’s health, development, and maltreatment.
Participants. One hundred ninety-four mothers who were adolescents (less than age 19) at delivery. Data were gathered when children were 4 to 5 years of age. Twenty-six percent of the children lived in 3-generation households, 39% had a history of maltreatment, and 32% of the mothers had depression scores in the clinical range.
Main Outcome Measures. Child behavioral problems were measured with the Child Behavior Checklist, completed by the mother, and child developmental status was assessed with the Battelle Developmental Inventory Screening Test, administered by research assistants.
Results. Multiple regression analyses revealed that children who had been reported for maltreatment and had mothers with depressive symptoms had more externalizing behavior problems, compared with children who experienced neither risk or only 1 risk. However, when residential status was considered, children with the greatest number of externalizing behavior problems were those who experienced both maltreatment and maternal depressive symptoms and lived in 3-generation households. Children who had been reported for maltreatment or had mothers with depressive symptoms were more likely to have internalizing problems, compared with children with neither risk. Residential status was not related to children’s internalizing behavior problems or cognitive development.
Conclusions. Living in a 3-generation household did not protect preschool children from the behavior problems associated with maltreatment and depression. In contrast, living in a 3-generation household was associated with more behavior problems among the highest risk group of children—those who had been maltreated and had mothers with symptoms of depression. Although 3-generation families may provide an important source of support and stability for adolescent mothers and their infants early in the parenting process, it may not be advisable to rely on 3-generation households as young mothers enter adulthood, particularly among those with a history of maltreatment or depression. Children with the fewest number of behavior problems were living with their mothers in their own household (often with the father), had not been maltreated, and had mothers with few symptoms of depression.
Key Words: adolescent mother • 3-generation • grandmother • maternal depression • maltreatment
Abbreviations: CPS, Child Protective Services • CBCL, Child Behavior Checklist
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This investigation targets young women who gave birth as adolescents and are transitioning to adulthood. The transition is marked by the emerging demands of adulthood and parenthood, making social support extremely important.12,17 In the past, much of the attention on parenting support has focused on the parents’ marital relationship. However, 79% of the mothers who give birth as adolescents are single.18 Rather than marrying and setting up independent households, most live in their family of origin and share caregiving with their mother (infant’s grandmother).19,20 Grandmothers are often viewed as providing support, nurturance, and sociological, financial, and legal stability.21 The importance of family support for young mothers has captured the attention of policymakers, as illustrated in the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, which requires adolescent mothers to live with a guardian to receive financial assistance.22
Despite the enthusiasm of policymakers regarding 3-generation households for adolescent mothers and their children, there are limited data on 3-generation households beyond the mother’s adolescent years.23 Grandmothers are often instrumental in the early phases of adolescent parenting, providing guidance and care for the child and enabling the mother to continue her education. Investigators have reported that children raised with both a mother and a grandmother in the household, rather than by a single parent, have better social adaptation during childhood,19,24 fewer behavior problems,25 less deviant behavior during adolescence,26 and, for children with low birth weight, better cognitive and health outcomes at 3 years of age.27 On the other hand, recent studies of 3-generation households have documented tensions related to role transitions and childrearing responsibilities involving young mothers and grandmothers.23,28–32
This investigation was designed to examine whether the 3-generation family living pattern is associated with fewer behavior problems and better development among preschool children of mothers who gave birth during their adolescence and whether the relationship varies among children who have experienced maltreatment and maternal depression. We tested 3 hypotheses. First, we hypothesized that children who lived in 3-generation households would have fewer behavior problems and better development than children living in independent households. Second, we hypothesized that children who had been maltreated or had a mother with depressive symptoms would be at increased risk for behavioral and developmental problems, particularly when the 2 risks co-occurred.33,34 Finally, we expected that grandmothers may be particularly important in the face of maltreatment and maternal depression. Therefore, we hypothesized that 3-generation households would protect children who had been maltreated or raised with a mother who reported depressive symptoms.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Children and mothers were recruited from 4 sites across the United States: the Northwest (N = 44), Midwest (N = 35), South (N = 69), and East (N = 46) and varied in their maltreatment status. In the northwestern site, all children had been reported to Child Protective Services (CPS). In the midwestern site, 17% of the children had been reported to CPS, and the remaining children were recruited from an agency serving high-risk families. In the southern site, children were identified at birth through a statewide High Priority Infant Program, based on medical and sociodemographic risk. At 4 years of age, 30% had been reported to CPS. In the eastern site, children were recruited from primary and specialty health care clinics serving low-income, urban communities, and by 4 years of age, 11% had been reported to CPS.
Procedure
After local institutional review board approval, a set of common measures and procedures was implemented across all study sites. Mothers participated in a 1.5- to 2-hour face-to-face interview assessing sociodemographics, family structure, maternal functioning, and their children’s behavior. A developmental screening test was administered to each child by a trained research assistant. Families were compensated financially for their participation.
Measures
Household composition data were gathered with the use of a Family Chart. For each member of the household, the interviewer recorded the names, ages, and relationship to the study child on the Family Chart.
Maternal depression was measured using the Center for Epidemiologic Studies-Depression Scale.36 This instrument includes 20 items comprising 6 major aspects of depression: depressed mood, feelings of guilt and worthlessness, feelings of helplessness and hopelessness, psychomotor retardation, loss of appetite, and sleep disturbance. Scores are summed to yield a total score with higher total scores indicating more depressive symptoms. Scores above 16 have been associated with clinically diagnosed depression. The internal consistency of the measure for this sample was 0.89.
The Family APGAR37 was used to assess the mothers’ perception of family functioning by examining her satisfaction with 5 parameters of family functioning: Adaptability, Partnership, Growth, Affection, and Resolve. The response options assess the frequency of feeling satisfied with each parameter on a 3-point scale ranging from 1 (hardly ever) to 3 (almost always). Values for the 5 items are summed producing a total score, with higher scores indicating a greater degree of satisfaction with family functioning. The internal consistency for the sample was 0.85.
The Adult-Adolescent Parenting Inventory38 assessed the parenting and child-rearing attitudes of the mothers in this study. The inventory includes 32 items grouped into 4 scales: Inappropriate Expectations, Lack of Empathy, Approval of Corporal Punishment, and Role Reversal. Responses were scored on a 5-point Likert scale that ranged from Strongly Agree to Strongly Disagree where higher scores on each scale reflected more appropriate behavior. In this study, we used the Lack of Empathy scale because it was the only scale in which the scores reflected deficiencies in parenting attitudes in comparison with general population norms.38 The internal consistency for the Lack of Empathy scale for the sample was 0.85.
A history of child maltreatment was defined as having at least 1 report made to CPS before the evaluation when the child was 4 or 5 years of age. We used reported cases, rather than substantiated cases, because previous evidence among children reported for maltreatment has shown that, as a group, children with unsubstantiated reports are just as likely to experience negative outcomes as are children with substantiated reports.39 Decisions regarding substantiation often result more from policy and resource considerations than from the facts of the case,40,41 leading child maltreatment researchers to recommend that the research community not depend solely on substantiated reports because of their limited representation and reduced power.39–45 State records were searched in each of the 4 sites.
Child developmental status was assessed using the Battelle Developmental Inventory Screening Test.46 Standard scores were computed for the cognitive development domain with a mean of 100 and a standard deviation of 15.
Child behavior problems were measured by the Child Behavior Checklist (CBCL), completed by the mother.47 The CBCL for 4- to 18-year-old children includes 113 items and respondents report on the frequency of each behavior over the past 6 months (0 = not true, 1 = somewhat or sometimes true, 2 = very true or often true). Scores were summed to yield externalizing and internalizing problem behavior scores. Raw scores were used in analyses to take advantage of the full variation in scores, as recommended by Achenbach.47 Norm referenced T scores were used to facilitate comparisons with other studies (mean: 50; standard deviation: 10). Higher scores on each subscale indicate more of that particular behavioral problem, and a T score of 60, which is 1 standard deviation above the mean, serves as the clinical cutoff.
Data Analysis
The hypotheses were evaluated by 3 multiple regression analyses using ordinary least squares. These analyses examined the main effects of grandmother presence, the child’s maltreatment history (maltreated or not), and maternal depressive symptoms (Center for Epidemiologic Studies-Depression Scale score) on the dependent variables of internalizing behavior problems, externalizing behavior problems, and cognitive performance. Study site, maternal education, household poverty (use of public assistance), and household size were included as covariates because they were empirically or conceptually related to children’s behavior or development.
To evaluate the hypotheses, we included three 2-way interaction terms, plus one 3-way interaction term. The first interaction term (the interaction between maltreatment and grandmother presence) examined the moderating effects of grandmother presence on the relationship between maltreatment and children’s behavior and development. The second (the interaction between maternal depressive symptoms and grandmother presence) examined the moderating effects of grandmother presence on the relationship between maternal depressive symptoms and children’s behavior and development. The third (the interaction between maternal depressive symptoms and maltreatment) examined the effects of both maternal depressive symptoms and maltreatment on children’s behavior and development. Finally, we included the 3-way interaction term among maltreatment, maternal depressive symptoms, and grandmother presence to determine whether the relationship between maltreatment and children’s behavior and development was differentially influenced by maternal depressive symptoms and grandmother presence. As recommended by Cohen and Cohen,48 we conducted follow-up analyses to identify the locus of interactions only when the interactions were significant in the overall analysis.
To reduce the possibility of committing type I errors (reporting differences when differences do not exist) when conducting multiple comparisons, many investigators use Bonferroni’s correction (
/k where = the significance level and k = number of tests performed). However, this procedure has been criticized because it is so stringent that it increases the likelihood of type II errors (failure to report differences when differences exist).49 To reduce the possibility of both type I and type II errors in multiple comparisons, we used a correction procedure developed by Holm50 that was recommended in a recent review.49 The Holm procedure uses an overall of 0.10 and calculates significance based on the number of tests performed and an ordering of the P values. For internalizing behavior, significance resulted at P < .04, for externalizing it was P < .02, and for cognitive it was P < .01.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
2 = 22.24; P < .01). The children’s scores on standardized measures of behavior and development indicate higher than expected rates of problems (Table 1). One third of the children (33%) obtained T scores in the clinical range on the Externalizing Scale of the Child Behavior Checklist, whereas in population-based studies, only 16% of children would be expected to obtain scores in the clinical range. The children’s scores on the Internalizing Scale of the Child Behavior Checklist were consistent with population-based norms. On the Battelle Developmental Inventory Screening Test, the children obtained scores that were approximately 1 standard deviation lower than the population mean.
In the regression analyses examining the effects of grandmother presence, maltreatment, and depression on children’s behavior and development, there were significant main effects of maltreatment and depression on both internalizing and externalizing behavior, but no main effects on cognitive development (Table 2). In contrast to the first hypothesis that living in a 3-generation household would be associated with fewer behavior problems and better cognitive development, children who lived in 3-generation households tended to have more internalizing and externalizing behavior problems. However, the finding did not reach significance when corrected for multiple comparisons.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
When grandmothers and adolescent mothers live together, the grandmother may provide more support for her daughter’s adolescent role (eg, remaining in school), than for her maternal role. For example, adolescent mothers who live in 3-generation households have been described as less nurturant during play17 and during feeding28 than adolescent mothers living independently, perhaps because when grandmothers are in the home, young mothers defer to them and do not gain the skills of parenting. Furstenberg and colleagues13 have shown that prolonged mother-grandmother coresidence (5 years or more and extending beyond adolescence) is associated with a lack of economic success among adolescent mothers and behavioral and academic problems among their children. One possible explanation for this finding is that the most competent mothers leave their family of origin to set up an independent household, leaving the least competent mothers in 3-generation households.
Little is known about the reasons that daughters leave their home of origin with their young children. For many young mothers, leaving home is a rite of passage that occurs when they have completed their education, formed a stable partnership, and are ready for the autonomy and independence associated with raising their child.12 The data from this investigation would support that view. Although many of the children may have lived in 3-generation households early in their lives, by their later preschool years only one-fourth were being raised in 3-generation households. This rate is consistent with the 18% 3-generation households reported by Luster and colleagues in their investigation of preschoolers born during their mothers’ adolescent years.51 Mothers who did not live in 3-generation households were more likely to be living with a male partner, compared with mothers who were in 3-generation households.
Although it is possible that mothers who perceived their children’s behavior to be problematic were more likely to remain in their family of origin, an alternative possibility is that the role confusion that has been described in 3-generation households4,17,52,53 may have contributed to the children’s behavior problems. However, because the data are cross-sectional, inferences can not be made regarding causality. In this investigation, as in others, there are many unanswered questions regarding adolescents’ living situations. Subsequent investigations should examine the impact of various living arrangements on adolescent functioning, the reasons adolescents live where they do, and the mechanisms that determine optimal living arrangements.
Support
Although young mothers need support from their family, the functioning level of the family may influence the quality of the support they can provide. For example, Kalil, Spencer, Spieker, and Gilchrist54 have shown that when adolescent mothers and grandmothers coresided in the context of poor family cohesion, the young mothers were likely to report depressive symptoms. In the current investigation, family satisfaction did not differ by residential status. However, the functioning level of the family should be considered in evaluating the merits of 3-generation households for young mothers and their children.
Social support is a central component in the adjustment of adolescent mothers.52,53 However, high levels of support may foster dependence and interfere with the development of the maturity and autonomy required for adulthood and parenthood.4,17,52 For example, Way and Leadbeater53 found that young mothers who received high levels of emotional support from their families were not more likely to achieve long-term educational attainment. One possible explanation is that when families are less supportive, young mothers may be driven toward educational achievement and success to prove themselves to their family or to gain the resources that will allow them to become independent. Another possibility is that social support is a multidimensional concept that changes as the developmental needs of young mothers change. During their teen years, young mothers may benefit from guidance and caregiving assistance, particularly as they complete the adolescent tasks of education and career preparation. However, as they mature into young adults, young mothers may benefit from support that promotes their autonomy and independence as responsible parents. As Spieker and Bensley17 have shown, preschool children of young mothers who gave birth during their adolescence, were more likely to be securely attached when their mothers had a supportive relationship with their own mother, but did not co-reside with her.
Maternal Depressive Symptoms and Maltreatmen
Approximately one third of the mothers reported depressive symptoms that were in the clinical range. These rates are slightly lower than reports from other studies of adolescent mothers, which tend to range from 40% to 69%,4–7 perhaps because most of the mothers in this investigation had entered adulthood and were beyond their adolescent years. Slightly more than one third (39%) of the children experienced maltreatment, primarily neglect. As expected, children who experienced maltreatment or maternal depression were more likely to have internalizing and externalizing behavior problems, even after adjusting for potential confounders. These findings are consistent with the literature that documents the negative consequences of maltreatment and maternal depression on children’s behavioral problems.55–60 However, our findings extend the literature on risk by documenting the exponential consequences of the 2 risks considered together. Consistent with the accumulation of risk model,33,34 the relationship between maltreatment and children’s externalizing behavior problems varied by their exposure to maternal depression. In other words, when mothers had few depressive symptoms, they may have been able to partially protect their children from the negative aspects of maltreatment. However, when both risks were present, children had high rates of externalizing behavior problems. Of course it is also possible that mothers of children with multiple externalizing problems experienced more depressive symptoms than mothers of children with few externalizing problems. Nevertheless, these findings highlight the importance of considering maternal depression in attempts to understand mechanisms linking contextual variables, such as maltreatment, with children’s behavior. The prevalence of depression in this sample also emphasizes the importance for pediatricians to screen for depressive symptoms among young mothers and to monitor the well-being of their children very closely.
The relation among maltreatment, maternal depression, and children’s externalizing behavior problems was further clarified with the introduction of grandmother coresidence to the model. Not only were grandmothers unable to protect children from the negative effects of maltreatment and maternal depression, but the situation was exacerbated when the children lived in a 3-generation household. In other words, children had more behavior problems when they lived in a 3-generation household. Although we can not be sure whether the most disturbed children and mothers remained in their family of origin or whether grandmothers contributed to the children’s problems, there is no evidence to support a protective effect of living in a 3-generation household.
Limitations
The adolescent parents in this investigation were selected based on risk criteria. This study should be replicated among other samples of young mothers to determine whether the findings generalize. In addition, at the time of the investigation the children were 4 to 5 years of age and most mothers were in their early 20s, although they had given birth as adolescents. Thus, the findings do not necessarily generalize to adolescent mothers when their children are infants.
Although we included a measure of family functioning, we did not include a specific measure of the quality of the mother-grandmother relationship. Grandmothers who attempt to help or support their daughters in parenting their children may undermine their daughters’ maternal role and growth toward autonomy. A measure of support would need to be sensitive to this issue for teens and young adults. Subsequent investigations into mother-grandmother living patterns should include measures of support that are sensitive to the developmental needs of teens and young adults and the quality of the mother-grandmother relationship.
Implications
This investigation has implications for theory, practice, and policy, and highlights the importance of considering the familial and psychological context surrounding the children of young parents. By the end of their preschool years, the most well-adjusted children were living in independent households with their mother (and often their father), had not experienced maltreatment, and did not have mothers with symptoms of depression. In attempts to help young families with multiple risks, policymakers and health care providers have often looked to grandmothers for support and stability. However, findings from this investigation showed that children confronted by multiple risks had more behavior problems when they lived in 3-generation households, compared with children who lived in independent households. Even in families with few risks, living in a 3-generation household did not protect preschoolers from behavior and developmental problems. Although the cross-sectional nature of the data do not allow us to determine the reasons behind the findings, they do suggest caution in relying on 3-generation families to protect children in difficult circumstances.
The type and quantity of support, and the balance between support and independence, may vary with the mother’s development and level of maturity. For example, early in their parenting role, young mothers may benefit from support for their adolescent role as they complete their education and prepare for the responsibilities associated with adulthood. As young mothers acquire these skills and transition into adulthood and a move away from their family of origin, they may benefit from support that promotes their autonomy and independence.
The evidence supporting the deleterious effect that maternal depression can have on children’s behavior has been well documented.60 Fortunately, the development of effective treatments for adolescents with depressive symptoms,61,62 often based on cognitive-behavioral therapy, offers hope for the reduction of maternal depressive symptoms among young mothers. Given the high rates of depression among adolescent mothers63 and the sensitivity of brief screening methods administered through paper and pencil tests,64,65 procedures for screening and referral (and perhaps treatment) could be introduced into sites that provide health care and education to pregnant and parenting teens.
Finally, despite the variability in the children’s behavior and development, they presented as a very vulnerable group. In comparison to population norms, the children were more likely to have externalizing behavior problems and to have low scores on a standardized assessment of cognitive development. Thus, they are at increased risk for the academic and emotional problems that have been described among children of adolescent parents. Alarming as this finding is, it is consistent with other reports of the vulnerability among children of adolescent mothers, particularly those with other risk factors12,13 and with the low scores in cognitive performance that often occur among preschool children from low-income communities, regardless of the mother’s age.66,67 Thus, there is a need to move away from stereotypes regarding children of adolescent parents and to examine theories regarding protective factors that can be built into intervention programs for pregnant and parenting teens and their children.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Reprint requests to (M.M.B.) Department of Pediatrics, University of Maryland School of Medicine, 655 W Lombard Street, Suite 311, Baltimore, MD 21201. E-mail: mblack{at}umaryland.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Maposa and L. SmithBattle Preliminary Reliability and Validity of the Grandparent Version of the Grandparent Support Scale for Teenage Mothers (GSSTM-G) Journal of Family Nursing, May 1, 2008; 14(2): 224 - 241. [Abstract] [PDF]
|
|
|
|
 |
|
 J. S. McCrae and R. P. Barth Using Cumulative Risk to Screen for Mental Health Problems in Child Welfare Research on Social Work Practice, March 1, 2008; 18(2): 144 - 159. [Abstract] [PDF]
|
|
|
|
 |
|
 L. SmithBattle Family legacies in shaping teen mothers' caregiving practices over 12 years. Qual Health Res, October 1, 2006; 16(8): 1129 - 1144. [Abstract] [PDF]
|
|
|
|
|
|
L. S. Sadler and D. A. Clemmens Ambivalent Grandmothers Raising Teen Daughters and Their Babies Journal of Family Nursing, May 1, 2004; 10(2): 211 - 231. [Abstract] [PDF]
|
|
|
|
 |
|
 C. Huaqing Qi and A. P. Kaiser Behavior Problems of Preschool Children From Low-Income Families: Review of the Literature Topics in Early Childhood Special Education, January 1, 2003; 23(4): 188 - 216. [Abstract] [PDF]
|
|
Read all P3Rs
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2002 American Academy of Pediatrics. All rights reserved. |
||
|
||
